Search results

Despite extensive data collection from many monitored bridges, this data often remains unprocessed and uninformative in its raw form.
Data-driven methods, on the other hand, bypass the physical model and directly analyze sensor data statistically to identify patterns linked to structural changes.
There is a common data environment that saves all relevant data for the model.
Here, the data is divided into training and test sets.
Although this error rate is already promising, it should be mentioned that these are just initial results using limited number of simulations with the potential for further reduction by increasing the training data volume in the future.

Monotonic mechanical properties at room temperature are 705 MPa ultimate strength, 385 MPa 0.2% proof strength, 25 % elongation, and 50 % reduction of area.
To avoid differences of the test data resulting from magnifications in observations, detectable size of the micro-cracks on each replica was kept to about 5 to 10 mum.
Corresponding to the microstructure barrier effect, this relation should be described in the mode as (9) Considering the scattered daI,eq/dN-ΔKeq data between the five specimens with effective crack information, a statistical modeling should be given.
First, the scattered data were checked with the method for determining an appropriate statistical distribution model under limited data [21], it verified that logarithm normal distribution is a an appropriate model for describing the daI,eq/dN data at a given ΔKeq level.
Second, random modeling of daI,eq/dN-ΔKeq relation is therefore reasonably established as (10) where daI,eq/dN|av is the average value of the daI,eq/dN data at a given ΔKeq level, daI,eq/dN|sd is the standard deviation of the daI,eq/dN data at a given ΔKeq level, ns is size of test paired daI,eq/dN -ΔKeq data, sr is the residual standard deviation of the logarithm of daI,eq/dN -ΔKeq relation (Eq. 9) fitting into the test data, and the subscripts P and C are respectively the scattered character related probability and the sample size related confidence. daI,eq/dN|av and daI,eq/dN|sd can be represented as (11) Material constants Aav, Bav, Dav, Asd, Bsd, and Dsd should be evaluated using a maximum likely method same to Ref. [22] and the measured results are respectively -10.3567, -0.220573, 0.0001, 0.655603, -0.121404, and 1.90695 with 20 paired daI,eq/dN -ΔKeq data.

To improve the accuracy of simulation, the initial ground stress field and rock mechanics parameters were adopted by considering the in-situ monitoring data.
Due to the application of support systems, the reduction of stresses is observed.
The trend almost agrees with monitoring data, which shows that the numerical model is effective to accurately evaluate the effects of faults and joint groups.
The deformations of Shuibuya powerhouse from numerical modeling almost agree with the in-situ measure data.
Conclusions (1) Considering the monitoring data and numerical modeling results, we conclude that the ground stress regime is mainly governed by self-weight gravity of rock mass.

The measurement robot samples data at appropriate places.
The ant terminates its traversal and outputs the parameter subset as a candidate for data reduction.
i αi-1 ai-1 di θi 1 0 0 0 θ1 2 88.373 0.9573 0 θ2 The test data are listed in table 3.
Table 3 Test data before optimized.
Table 5 Test data after ACO is applied.

Based on a 24-yr hindcast wave data obtained from WW3 (WAVEWATCH-III) wave model forcing by CCMP (Cross-Calibrated, Multi-Platform) wind data, this study will present the wave energy characteristic of the South China Sea, such as the seasonal characteristic of wave power density, occurrence of energy level, the long-term trend of wave power density.
We hope this study can provide reference for the evaluation of wave energy. 4.1 Data The data used in this study is a 24-yr hindcast wave data.
Then verify the precision of the simulated wave data by using the Korean and Japanese buoy data on the wave.
Contrasting the simulated wave data with Korean and Japanese buoy data, we found the simulated wave data is trustworthy [6]. 4.2 Seasonal characteristic of wave power density In January (Fig.2a, represent as winter), the South China Sea wave power density is the largest of the four seasons, of about 12-27 kW/m in most area of the South China Sea.
In this paper, we make statistics of the occurrence of wave power density greater than 2kW/m using a 24-yr 3-hourly data, as shown in Fig.3.

In the same temperature window of 1300-2000°C, the RT Hall data show that both hole carrier density and hole mobility increase with increasing annealing time and annealing temperature.
Without a comparison with the C-V data and without a critical analysis of the Hall mobility data, the bare analysis of the Hall carrier density data would bring to the conclusion of an increased acceptor density with increasing annealing temperature and increasing annealing time.
The Hall mobility curves are the ratio of the correspondent Hall coefficient data to the sheet resistance ones.
Drift carrier density curves were derived from the Hall coefficient (RH) data by using the temperature-dependent Hall scattering factor rH(T) of [6,7].
From these data, the percentages of the ion-implanted Al electrical activation and of the acceptor compensation have been computed.

The observed reduction in the hardness could be attributed to the samples brittleness.
By applying equation (1) the flexural strength (σf) was determined [16]: σf = 2Fπd h (1) Thermal conductivity coefficient was calculated to the data that measurement by using the lee's disk {manufacture by Griffin and George/England}, thermal conductivity coefficient was calculated by using the following equations [17]: K [TB-TA/ds]=e[TA+2/r[dA+dS/4]TA+1/2r(dSTB) (2) H=IV=πr²e(TA+TB)+2πre[dATA+(1/2)dS(TA+TB)+dBTB+dCTC] (3) Where K is the thermal conductivity coefficient, e represents the amount of thermal energy passing through a unit area per second disk material, H represents the thermal energy passing through the heating coil unit of time, d is the thickness of the disks (mm), r is the radius of the disk (mm), dS is the thickness of the sample (mm), and T is the

Thirdly, information was gathered for machine learning models by extracting data from publication of Ruitao Peng et al[13].
In the present work, 80% of data is selected for the training while 20% of data is for the testing of ML models.
The ML models randomly select the data[33], [34].
Using the Python Pandas module, data was imported and processed from Microsoft Excel in the first stage.
The study is limited, though, mainly because machine learning models rely on the availability and quality of data.

The experimental data showed: (1) with the increasing of concentration and time for fumigation, SO2 had a great adverse effect on photosynthesis in soybean.
The reduction of gs indicated plants could react quickly against the adverse regime.

The stable stress distribution was obtained as Fig.4,The maximum stress was shown at the inner surface where flange valve intersect.The result agreed with the expection before calculation because it is a geometric discontinuous location.The stress verification line was drawn here and the datum was shown at table 1.
Table 3 stress analysis result under different inner pressure Inner presure (MPa) SⅡ(MPa) SⅣ(MPa) Result 17 237.1>205.5 355.6<411 No allow 16 226.4>205.5 348.2<411 No allow 15 214.9>205.5 337.5<411 No allow 14 202.4<205.5 323.5<411 allow 13 189<205.5 304.4<411 allow From the table 3 we found that follow the reduction of inner pressure the value at stress analysis descend accordingly.